Security audit

代码解释工具专业版

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly a code-documentation helper, but it asks for broad command/write authority and recommends unattended automation without enough safeguards.

Review this skill before installing. It appears intended for local code analysis and documentation generation, not theft or destruction, but only use it on repositories where project-wide reading and generated-file writes are acceptable. Avoid enabling cron, Git hooks, auto-update, or search-index features unless outputs are limited to reviewed docs/wiki paths and sensitive directories are excluded.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Vague Triggers

High

Confidence: 91% confidence
Finding: An overly broad and mismatched trigger condition can cause the skill to activate for unrelated tasks, increasing the chance that command execution, file reads, or file writes occur in contexts the user did not intend. In this skill, that risk is amplified because the declared toolset includes exec, write, glob, and grep, so unintended invocation could lead to unnecessary workspace modification or broad codebase inspection.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill describes generating files and running shell/Python commands that modify the workspace, but it does not clearly warn users that local files will be created or overwritten. In an agent setting, this can mislead users into invoking what appears to be an analysis tool while actually performing state-changing operations across the repository.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The documentation recommends unattended scheduled runs and Git hook automation that can repeatedly modify documentation and repository contents without an interactive safety check. In practice, this increases the chance of persistent unintended changes, CI churn, noisy commits, or accidental inclusion of sensitive derived content in version control.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The documented skill includes code that automatically creates directories and writes multiple documentation files (for example via os.makedirs(...) and open(..., 'w')) across a project-derived output path, but the surrounding markdown does not clearly warn users that running it will modify the filesystem. In an agent-skill context, silent write behavior is security-relevant because users may expect analysis or explanation only, while the tool can create or overwrite files in bulk.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.